linux-mm.kvack.org archive mirror
 help / color / mirror / Atom feed
* Re: [syzbot] [btrfs?] possible deadlock in __btrfs_release_delayed_node (2)
       [not found] <000000000000c826b605f748e7f2@google.com>
@ 2023-03-20  7:43 ` Hillf Danton
  0 siblings, 0 replies; only message in thread
From: Hillf Danton @ 2023-03-20  7:43 UTC (permalink / raw)
  To: Michal Hocko
  Cc: Tetsuo Handa, Linus Torvalds, linux-mm, linux-kernel, syzbot,
	syzkaller-bugs

On Sun, 19 Mar 2023 16:09:41 -0700
> syzbot found the following issue on:
> 
> HEAD commit:    9c1bec9c0b08 Merge tag 'linux-kselftest-fixes-6.3-rc3' of ..
> git tree:       upstream
> console output: https://syzkaller.appspot.com/x/log.txt?x=163b8aecc80000
> kernel config:  https://syzkaller.appspot.com/x/.config?x=dbab9019ad6fc418
> dashboard link: https://syzkaller.appspot.com/bug?extid=3ae9507d4e2431b56ff4
> compiler:       Debian clang version 15.0.7, GNU ld (GNU Binutils for Debian) 2.35.2
> 
> Unfortunately, I don't have any reproducer for this issue yet.
> 
> Downloadable assets:
> disk image: https://storage.googleapis.com/syzbot-assets/adcd3c9a01d5/disk-9c1bec9c.raw.xz
> vmlinux: https://storage.googleapis.com/syzbot-assets/3f5bb295bb37/vmlinux-9c1bec9c.xz
> kernel image: https://storage.googleapis.com/syzbot-assets/029aafd259b9/bzImage-9c1bec9c.xz
> 
> IMPORTANT: if you fix the issue, please add the following tag to the commit:
> Reported-by: syzbot+3ae9507d4e2431b56ff4@syzkaller.appspotmail.com
> 
> ======================================================
> WARNING: possible circular locking dependency detected
> 6.3.0-rc2-syzkaller-00050-g9c1bec9c0b08 #0 Not tainted
> ------------------------------------------------------
> syz-executor.1/26441 is trying to acquire lock:
> ffff8880286e5800 (&delayed_node->mutex){+.+.}-{3:3}, at: __btrfs_release_delayed_node+0x9a/0xaa0 fs/btrfs/delayed-inode.c:256
> 
> but task is already holding lock:
> ffff8880592c24d8 (btrfs-root-01#2){++++}-{3:3}, at: __btrfs_tree_lock+0x3c/0x2a0 fs/btrfs/locking.c:197
> 
> which lock already depends on the new lock.
> 
> 
> the existing dependency chain (in reverse order) is:
> 
> -> #3 (btrfs-root-01#2){++++}-{3:3}:
>        lock_acquire+0x1e1/0x520 kernel/locking/lockdep.c:5669
>        down_read_nested+0x40/0x60 kernel/locking/rwsem.c:1645
>        __btrfs_tree_read_lock+0x2f/0x220 fs/btrfs/locking.c:135
>        btrfs_tree_read_lock fs/btrfs/locking.c:141 [inline]
>        btrfs_read_lock_root_node+0x292/0x3c0 fs/btrfs/locking.c:280
>        btrfs_search_slot_get_root fs/btrfs/ctree.c:1774 [inline]
>        btrfs_search_slot+0x571/0x2f70 fs/btrfs/ctree.c:2096
>        btrfs_insert_empty_items fs/btrfs/ctree.c:4202 [inline]
>        btrfs_insert_empty_item fs/btrfs/ctree.h:646 [inline]
>        btrfs_insert_item+0x198/0x3a0 fs/btrfs/ctree.c:4231
>        create_pending_snapshot+0xdef/0x28c0 fs/btrfs/transaction.c:1783
>        create_pending_snapshots+0x195/0x1d0 fs/btrfs/transaction.c:1894
>        btrfs_commit_transaction+0x1304/0x3440 fs/btrfs/transaction.c:2351
>        create_snapshot+0x4a5/0x7e0 fs/btrfs/ioctl.c:844
>        btrfs_mksubvol+0x5d0/0x750 fs/btrfs/ioctl.c:994
>        btrfs_mksnapshot+0xb5/0xf0 fs/btrfs/ioctl.c:1040
>        __btrfs_ioctl_snap_create+0x338/0x450 fs/btrfs/ioctl.c:1293
>        btrfs_ioctl_snap_create+0x136/0x190 fs/btrfs/ioctl.c:1320
>        btrfs_ioctl+0xbbc/0xd40
>        vfs_ioctl fs/ioctl.c:51 [inline]
>        __do_sys_ioctl fs/ioctl.c:870 [inline]
>        __se_sys_ioctl+0xf1/0x160 fs/ioctl.c:856
>        do_syscall_x64 arch/x86/entry/common.c:50 [inline]
>        do_syscall_64+0x41/0xc0 arch/x86/entry/common.c:80
>        entry_SYSCALL_64_after_hwframe+0x63/0xcd
> 
> -> #2 (btrfs-tree-01/7){+.+.}-{3:3}:
>        lock_acquire+0x1e1/0x520 kernel/locking/lockdep.c:5669
>        down_write_nested+0x3d/0x50 kernel/locking/rwsem.c:1689
>        __btrfs_tree_lock+0x3c/0x2a0 fs/btrfs/locking.c:197
>        btrfs_init_new_buffer fs/btrfs/extent-tree.c:4840 [inline]
>        btrfs_alloc_tree_block+0x515/0x1800 fs/btrfs/extent-tree.c:4918
>        btrfs_copy_root+0x345/0xcf0 fs/btrfs/ctree.c:327
>        create_pending_snapshot+0xcc7/0x28c0 fs/btrfs/transaction.c:1768
>        create_pending_snapshots+0x195/0x1d0 fs/btrfs/transaction.c:1894
>        btrfs_commit_transaction+0x1304/0x3440 fs/btrfs/transaction.c:2351
>        create_snapshot+0x4a5/0x7e0 fs/btrfs/ioctl.c:844
>        btrfs_mksubvol+0x5d0/0x750 fs/btrfs/ioctl.c:994
>        btrfs_mksnapshot+0xb5/0xf0 fs/btrfs/ioctl.c:1040
>        __btrfs_ioctl_snap_create+0x338/0x450 fs/btrfs/ioctl.c:1293
>        btrfs_ioctl_snap_create+0x136/0x190 fs/btrfs/ioctl.c:1320
>        btrfs_ioctl+0xbbc/0xd40
>        vfs_ioctl fs/ioctl.c:51 [inline]
>        __do_sys_ioctl fs/ioctl.c:870 [inline]
>        __se_sys_ioctl+0xf1/0x160 fs/ioctl.c:856
>        do_syscall_x64 arch/x86/entry/common.c:50 [inline]
>        do_syscall_64+0x41/0xc0 arch/x86/entry/common.c:80
>        entry_SYSCALL_64_after_hwframe+0x63/0xcd
> 
> -> #1 (btrfs-tree-01){++++}-{3:3}:
>        lock_acquire+0x1e1/0x520 kernel/locking/lockdep.c:5669
>        down_read_nested+0x40/0x60 kernel/locking/rwsem.c:1645
>        __btrfs_tree_read_lock+0x2f/0x220 fs/btrfs/locking.c:135
>        btrfs_tree_read_lock fs/btrfs/locking.c:141 [inline]
>        btrfs_read_lock_root_node+0x292/0x3c0 fs/btrfs/locking.c:280
>        btrfs_search_slot_get_root fs/btrfs/ctree.c:1774 [inline]
>        btrfs_search_slot+0x571/0x2f70 fs/btrfs/ctree.c:2096
>        btrfs_insert_empty_items+0x9c/0x180 fs/btrfs/ctree.c:4202
>        btrfs_insert_delayed_item fs/btrfs/delayed-inode.c:746 [inline]
>        btrfs_insert_delayed_items fs/btrfs/delayed-inode.c:824 [inline]
>        __btrfs_commit_inode_delayed_items+0xd53/0x2400 fs/btrfs/delayed-inode.c:1111
>        __btrfs_run_delayed_items+0x1db/0x430 fs/btrfs/delayed-inode.c:1153
>        flush_space+0x26d/0xe30 fs/btrfs/space-info.c:729
>        btrfs_async_reclaim_metadata_space+0x106/0x350 fs/btrfs/space-info.c:1087
>        process_one_work+0x8a0/0x10e0 kernel/workqueue.c:2390
>        worker_thread+0xa63/0x1210 kernel/workqueue.c:2537
>        kthread+0x270/0x300 kernel/kthread.c:376
>        ret_from_fork+0x1f/0x30 arch/x86/entry/entry_64.S:308
> 
> -> #0 (&delayed_node->mutex){+.+.}-{3:3}:
>        check_prev_add kernel/locking/lockdep.c:3098 [inline]
>        check_prevs_add kernel/locking/lockdep.c:3217 [inline]
>        validate_chain+0x166b/0x58e0 kernel/locking/lockdep.c:3832
>        __lock_acquire+0x125b/0x1f80 kernel/locking/lockdep.c:5056
>        lock_acquire+0x1e1/0x520 kernel/locking/lockdep.c:5669
>        __mutex_lock_common+0x1d8/0x2530 kernel/locking/mutex.c:603
>        __mutex_lock kernel/locking/mutex.c:747 [inline]
>        mutex_lock_nested+0x1b/0x20 kernel/locking/mutex.c:799
>        __btrfs_release_delayed_node+0x9a/0xaa0 fs/btrfs/delayed-inode.c:256
>        btrfs_evict_inode+0x6f4/0x1090 fs/btrfs/inode.c:5398
>        evict+0x2a4/0x620 fs/inode.c:665
>        dispose_list fs/inode.c:698 [inline]
>        prune_icache_sb+0x239/0x2e0 fs/inode.c:897
>        super_cache_scan+0x376/0x480 fs/super.c:106
>        do_shrink_slab+0x544/0xdb0 mm/vmscan.c:853
>        shrink_slab_memcg mm/vmscan.c:922 [inline]
>        shrink_slab+0x578/0x8c0 mm/vmscan.c:1001
>        shrink_node_memcgs mm/vmscan.c:6439 [inline]
>        shrink_node+0x1143/0x2730 mm/vmscan.c:6473
>        shrink_zones mm/vmscan.c:6711 [inline]
>        do_try_to_free_pages+0x67e/0x1900 mm/vmscan.c:6773
>        try_to_free_mem_cgroup_pages+0x455/0xa50 mm/vmscan.c:7088
>        reclaim_high+0x1e5/0x270 mm/memcontrol.c:2403
>        mem_cgroup_handle_over_high+0x14a/0x2b0 mm/memcontrol.c:2588

Hi Michal

3) Anything other than defer handle_over_high to workqueue available in
order to avoid deadlock like this one, given GFP_NOFS?

>        try_charge_memcg+0x13b5/0x16d0 mm/memcontrol.c:2826
>        try_charge mm/memcontrol.c:2837 [inline]
>        charge_memcg+0x11a/0x3f0 mm/memcontrol.c:6960
>        __mem_cgroup_charge+0x27/0x80 mm/memcontrol.c:6981
>        mem_cgroup_charge include/linux/memcontrol.h:678 [inline]
>        __filemap_add_folio+0xe78/0x1b50 mm/filemap.c:857
>        filemap_add_folio+0x121/0x580 mm/filemap.c:939
>        __filemap_get_folio+0x7d5/0xe50 mm/filemap.c:1981

2) Add folio to pagecache.

>        pagecache_get_page+0x2c/0x240 mm/folio-compat.c:99
>        find_or_create_page include/linux/pagemap.h:632 [inline]

1) Allocate page with gfp = GFP_NOFS|__GFP_NOFAIL if no page in pagecache.

>        alloc_extent_buffer+0x252/0xff0 fs/btrfs/extent_io.c:4044
>        btrfs_init_new_buffer fs/btrfs/extent-tree.c:4799 [inline]
>        btrfs_alloc_tree_block+0x2a0/0x1800 fs/btrfs/extent-tree.c:4918
>        __btrfs_cow_block+0x470/0x1830 fs/btrfs/ctree.c:541
>        btrfs_cow_block+0x403/0x780 fs/btrfs/ctree.c:696
>        btrfs_search_slot+0xc89/0x2f70 fs/btrfs/ctree.c:2136
>        del_balance_item fs/btrfs/volumes.c:3502 [inline]
>        reset_balance_state+0x1e5/0x3a0 fs/btrfs/volumes.c:3577
>        btrfs_balance+0xf17/0x1120 fs/btrfs/volumes.c:4415
>        btrfs_ioctl_balance+0x493/0x7c0 fs/btrfs/ioctl.c:3592
>        vfs_ioctl fs/ioctl.c:51 [inline]
>        __do_sys_ioctl fs/ioctl.c:870 [inline]
>        __se_sys_ioctl+0xf1/0x160 fs/ioctl.c:856
>        do_syscall_x64 arch/x86/entry/common.c:50 [inline]
>        do_syscall_64+0x41/0xc0 arch/x86/entry/common.c:80
>        entry_SYSCALL_64_after_hwframe+0x63/0xcd
> 
> other info that might help us debug this:
> 
> Chain exists of:
>   &delayed_node->mutex --> btrfs-tree-01/7 --> btrfs-root-01#2
> 
>  Possible unsafe locking scenario:
> 
>        CPU0                    CPU1
>        ----                    ----
>   lock(btrfs-root-01#2);
>                                lock(btrfs-tree-01/7);
>                                lock(btrfs-root-01#2);
>   lock(&delayed_node->mutex);
> 
>  *** DEADLOCK ***
> 
> 8 locks held by syz-executor.1/26441:
>  #0: ffff88807eb2e460 (sb_writers#14){.+.+}-{0:0}, at: mnt_want_write_file+0x5e/0x1f0 fs/namespace.c:438
>  #1: ffff88807c1cd440 (&fs_info->balance_mutex){+.+.}-{3:3}, at: btrfs_balance+0xbeb/0x1120 fs/btrfs/volumes.c:4381
>  #2: ffff88807eb2e650 (sb_internal#2){.+.+}-{0:0}, at: del_balance_item fs/btrfs/volumes.c:3492 [inline]
>  #2: ffff88807eb2e650 (sb_internal#2){.+.+}-{0:0}, at: reset_balance_state+0x12e/0x3a0 fs/btrfs/volumes.c:3577
>  #3: ffff88807c1ce390 (btrfs_trans_num_writers){++++}-{0:0}, at: spin_unlock include/linux/spinlock.h:390 [inline]
>  #3: ffff88807c1ce390 (btrfs_trans_num_writers){++++}-{0:0}, at: join_transaction+0xbfd/0xe80 fs/btrfs/transaction.c:287
>  #4: ffff88807c1ce3b8 (btrfs_trans_num_extwriters){++++}-{0:0}, at: join_transaction+0xc28/0xe80 fs/btrfs/transaction.c:288
>  #5: ffff8880592c24d8 (btrfs-root-01#2){++++}-{3:3}, at: __btrfs_tree_lock+0x3c/0x2a0 fs/btrfs/locking.c:197
>  #6: ffffffff8cfd4690 (shrinker_rwsem){++++}-{3:3}, at: shrink_slab_memcg mm/vmscan.c:895 [inline]
>  #6: ffffffff8cfd4690 (shrinker_rwsem){++++}-{3:3}, at: shrink_slab+0x2dc/0x8c0 mm/vmscan.c:1001
>  #7: ffff88807eb2e0e0 (&type->s_umount_key#77){++++}-{3:3}, at: trylock_super fs/super.c:414 [inline]
>  #7: ffff88807eb2e0e0 (&type->s_umount_key#77){++++}-{3:3}, at: super_cache_scan+0x77/0x480 fs/super.c:79


^ permalink raw reply	[flat|nested] only message in thread

only message in thread, other threads:[~2023-03-20  7:43 UTC | newest]

Thread overview: (only message) (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
     [not found] <000000000000c826b605f748e7f2@google.com>
2023-03-20  7:43 ` [syzbot] [btrfs?] possible deadlock in __btrfs_release_delayed_node (2) Hillf Danton

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).